Considerations for the Return of Genomic Results @HeidiRehm Heidi L. - - PowerPoint PPT Presentation
Considerations for the Return of Genomic Results @HeidiRehm Heidi L. Rehm, PhD, FACMG Medical Director, Broad Institute Clinical Research Sequencing Platform Associate Professor of Pathology, Brigham and Womens Hospital and Harvard Medical
Heidi L. Rehm, PhD, FACMG
Medical Director, Broad Institute Clinical Research Sequencing Platform Associate Professor of Pathology, Brigham and Women’s Hospital and Harvard Medical School
@HeidiRehm
▫ SNVs challenging in homologous regions ▫ InDels often challenging
Consider making some data available separately as unconfirmed research results requiring CLIA confirmation
(e.g. PMS2, SMN1/2, VWF, STRC)
Courtesy: Diana Mandelker and Birgit Funke
Require Sanger Assume TP Assume FP
Novel null, 34% Reported null, 27% Reported missense, 39% ~5 million variants Technical exclusions
HGMD
<5% <1%
Gene exclusions Variant exclusions ~100‐200 variants ~10 variants for manual review
Medical exome Novel LOF ClinVar *In non‐diagnostic testing, 92% of variants reported as pathogenic in HGMD had insufficient evidence to support the claim.
~5000 genes
~2% (0‐7) of filtered variants reported in MedSeq Several hours of review per genome Pathogenic Likely Pathogenic VUS – Favor Pathogenic Variant Type Average Review Time Variant with literature 90 min Variant with no literature 26 min
26% (97,422/377,075) of variants have ≥2 submitters in ClinVar 17% (16,631/97,422) are interpreted differently
NEJM May 27th, 2015
3.6% medically significant (P/LP vs VUS/LB/B) 1.7% medically significant among clinical lab submissions 11% (12,895/118,169) of variants have ≥2 submitters in ClinVar 17% (2229/12,895) are interpreted differently
Discrepancy Resolution Process
Discrepancy Resolution Process
Harrison et al. Genet Med, in press Steven Harrison
Ambry Chicago GeneDx LMM
~4% of cases per year received medium or high alerts
Genet Med. Apr 2012 PMID: 22481129
MedSeq Genome Reanalysis
22% (22/100) Participants Received New or Reclassified Variants Both, 3 Reclassified Finding(s), 10 Previously Unreported Finding(s), 9
Expert Panel interpretations sometimes change as well
Variant Reclassification Over 12 Years at the Laboratory for Molecular Medicine
Reports, Structured Variant Data and Variant Updates Returned Via EHR (GeneInsight Clinic)
▫ ACMG59 as starting point ▫ Pathogenic as starting point
▫ Enable participants to share their raw data broadly Array genotypes, BAMs, VCFs List of annotated novel, rare or suspicious variants ▫ Allow access to data when clinical context raises the prior probability of disease CLIA confirmations and interpretations can be ordered as needed ▫ Enable other studies that can delve deeper into the significance of these variants
Are null variants an established mechanism of disease? How frequently are predicted null variants found in the gene in large population databases? Het vs hom? What is the constraint score in ExAC? Are there other known pathogenic variants in the exon? Also check ExAC for nulls in that exon. Is nonsense‐mediated decay predicted? Are there predicted null variants reported 3' (and 5’) to the variant? Is the exon alternatively spliced?
Hereditary Cancer Cardiovascular Disease Inborn Errors of Metabolism Neuro developmental Hearing Loss RASopathies EP RASopathies EP Other CDH1 EP PTEN EP TP53 EP Cardiomyopathy Cardiomyopathy EP KCNQ1 EP FH EP PAH EP Mito EP Brain Malformations EP Hearing Loss EP Application approved by ClinGen for 3 star submissions to ClinVar Planning to apply to ClinGen for 3 star submission level MODY EP ENIGMA EP CFTR EP CFTR EP InSiGHT EP PharmGKB PharmGKB
FAO EP Rett/Angelman‐ like Disorders
9318 expert reviewed variants in ClinVar (2.5%)
Somatic & Germline Cancer Curation Group Aminoacid‐
▫ Single submitter criteria provided (1 star) ▫ Experienced clinical lab subjective – opinion of physicians and clinical lab peers objective measures – volume of submissions in a disease area (data from ClinVar Miner)
50000 100000 150000 200000 250000 0 star 1 star w/ conflict 1 star single submitter 2 star 3/4 star
hu025CEA (Heidi Rehm) ‐ GET‐Evidence variant report– PGP Project
http://evidence.pgp‐hms.org/genomes
Genome report
Variant Clinical Importance Impact Allele freq Summary
APOE‐C130R
High Well‐established pathogenic 14% This is generally known as the ApoE4 allele of ApoE and is associated with increased risk of Alzheimer's. 20‐25% of individuals are heterozygous for this variant, and 1‐2% are homozygous. Data from Khachaturian et al. suggests an average 7% of all individuals developed Alzheimer's by the age of 80; when this is split by ApoE4 status: 10% of ApoE4 heterozygotes (3% increased attributable risk), 40% of ApoE4 homozygotes (33% increased attributable risk), and 5% of non‐carriers (2% decreased attributable risk). Notably, their model suggests 70‐75% of people would eventually develop Alzheimer's by the age of 100 regardless of ApoE4 genotype (and 25‐30% are resistant, regardless of genotype), but that ApoE4 variants shift the disease onset to occur significantly earlier (4 years earlier for heterozygous carriers, 13 years for homozygotes). Complex/Other, Heterozygous
NOD2‐R702W
Low Likely pathogenic 3.30% NOD2 encodes a protein involved in bacterial recognition. This variant is associated with Crohn's disease in European populations, but not in Korean or Japanese groups. Complex/Other, Heterozygous
MBL2‐R52C
Low Likely pathogenic 4.90% This variant is associated with mannose binding protein deficiency which leads to impaired complement system immune response to mannose‐rich pathogens. Patients homozygous for this allele or compound heterozygous are likely to have increased susceptibility to infection, but Hellemann et al. report heterosis for intensive care outcomes in heterozygous subjects. The wild‐type version of this gene is known as variant allele A, while this is called variant allele D. See G54D (variant B) and G57E (variant C). Recessive, Carrier (Heterozygous)
APOA5‐S19W
Low Likely pathogenic 6.50% This variant, also known as APOA5*3, is associated with higher plasma triglyceride concentrations but no significant correlation with coronary artery disease itself has been found. Unknown, Heterozygous
MTRR‐I49M
Low Likely pathogenic 45% This common variant (HapMap allele frequency of 31.3%) in a protein involved in folate (B9) and cobalamin (B12) metabolism and is often reported as "MTRR I22M" (an alternative transcript position). Mothers homozygous for this variant are associated with having around a increased chance of a child with Down syndrome (risk of 0.4%, average risk in population is 0.25%). Notably, age plays a far larger role in the rate of Down syndrome (risk is 4.5% for a mother 45‐years‐of‐age), and it is unknown how this variant may combine with the effect of age. There are conflicting reports associating this variant with incidence of neural tube defects, possibly when combined with MTHFR A222V. Recessive, Carrier (Heterozygous)
CD40LG‐G219R
Low Uncertain pathogenic 1.10% Study of a single family with X‐linked immunodeficiency implicated this variant as causal when combined with XIAP‐G466X. The authors' hypothesis is that either variant alone has much less effect, if
Recessive, Carrier (Heterozygous)
Insufficiently evaluated variants (3319 variants)
Variant Prioritization score Allele freq Num of articles Zygosity and Prioritization Score Reasons
MC2R‐S74I
5 0.02%
NEFL‐S472Shift
4 ?
RSPH4A‐W607Shift
4 ?
TTN‐E190Shift
4 0.86%
XDH‐R1296W
4 1.30%
CEP290‐E277Q
4 1.40%
DPYD‐S534N
4 1.60% 2 Carrier (Heterozygous). Has unevaluated web hits, Polyphen 2: 0.996 (probably damaging), Testable gene in GeneTests
CEP290‐K838E
4 3.20%
LAMC2‐D247E
4 3.40% 1
F5‐P1404S
4 3.70%
COL11A2‐P1316T
4 4.60%
IL23R‐R381Q
4 4.80% 3
ATP8B1‐R952Q
4 8.30%
RDH12‐R161Q
4 12%
SPG7‐R688Q
4 14% 3 Carrier (Heterozygous). In PharmGKB, Polyphen 2: 0.203 (possibly damaging), Testable gene in GeneTests with associated GeneReview
SPG7‐T503A
4 14% 1
DLL3‐F172C
4 15%
MSH6‐G39E
4 18% 3
ATP7A‐V767L
4 25% 1
GPR98‐Y2232C
4 32%
NHLRC1‐P111L
4 34%
Variant 56 of 3319: MYH7‐R1500W
3 ?
Case Source Year Dx Age Dx Segregatio n Clinical hx; Family 1 Karkkainen 2004 DCM 55 yr not tested mother heart failure/car several family members d 2 Jerosch‐Herold 2008 DCM 56 yr 2 mother symptoms at 95 ( consistent with DCM) and 73 of HF) and 63 yr (dysp 3 Merlo 2013 DCM ? ? 4 Hazebroek 2015 DCM ? (>18 yr) ? 5 LMM 2013 DCM/LVNC 32 yr no fam hx past hx of IV drug use; eje improved 6 GeneDx 20?? DCM <13 yr homozygous 7 GeneDx 20?? DCM 8 GeneDx 20?? DCM 9 Invitae 2015 DCM/LVNC 34 yr Hispanic no fam hx peripartum cardiomyopa hypothyroidism; heart ha thyroid treatment but sti list 10 Geisinger 2016
Director of Policy & Education Geisinger Health System wafaucett@geisinger.edu
Faucett WA & Davis FD, 2016, Appl Trans Genom
Faucett WA & Davis FD, 2016, Appl Trans Genom
evaluate & return
actionable genes
1. Disclose nature of results 2. Schedule follow‐up (detailed disclosure, clinical eval) 3. Encourage family communication of results
available for in‐person consults 3 days/week
Follow‐up Status Positive Results % Clinical Genomics 245 45% PCP or specialist 73 13% Declined immediate follow‐up 134 25% Lost to follow‐up 38 7% Deceased/Withdrawn 9 2% In Process 45 8% Total 544 100%
members…if you find something you can nip in the bud, it’s not nearly as expensive”
potential of having it themselves have been very laid back about it actually”
support
Detailed discussion of results by genetics clinician Targeted discussion of results by ordering clinician
Per clinician & participant preferences Condition associated with increased risk of adverse psychological impact? Is patient a minor who had an adult‐onset condition identified?
No No Yes Yes
Disclosure of results via phone, patient portal, or written material
Per clinician & participant preferences
Pathogenic/Likely Pathogenic Variant Result
ACMG SFv2.0 gene recommendations completed
Disclosure of Results by ordering clinician (via phone
priority referral
33
34
V 3.0 ACMG 2.0 & PGX Uninterpreted Data V 2.0 ACMG 2.0 Pharmacogenomics V 1.0 ACMG 2.0
Confirm interest with each version